Implications of Law decisions1 Address to Northern Ireland Personal Injury Bar’s Inaugural Conference, County Down Lord Neuberger, President of the Supreme Court 13 May 2017

1. In the Fairchild case2, which I shall discuss later, Lord Nicholls of Birkenhead described legal concepts determining the appropriate scope of liability in tort, including duty of care, , proximity, and remoteness, as “afflicted with linguistic ambiguity”. And, as the equally formidable Professor Jane Stapleton has written, the legal reasoning in judgments in tort cases is often obscure, so that it is difficult to distil a coherent body of principles3.

2. I would suggest (not for the first time4) that this is unsurprising as almost all aspects of tort law, above all , are based on policy. Of course, almost all legal principles can be said to be ultimately based on policy, but, given the very broad area which tort (in particular negligence) covers, and given the infinite variety of human life, any attempt to identify or distil clear principles in such an area is fraught with problems. Indeed, as the great Sir Frederick Pollock pointed out at the end of the 19th century, it is not even very easy to define what a tort is. As he wrote, “nothing seems easier than to [give] examples of , … [b]ut we shall have no such easy task if we are required to answer the question, What is a tort?”5 He went on to ask whether “we shall discover any general principles [of tort law] at all”, and said that he considered would be unlikely to happen6.

3. Analysis of tort cases appears to demonstrate a notable degree of disarray and a marked lack of reliable principle. Thus, what appears to be principle turns out, on examination, to be policy. A good example is to be found in the House of Lords Caparo decision7, which is part of the saga which started with Anns8 in 1977 (indeed, some might say it started with Donoghue

1 I am most grateful to Charlotte Gilmartin for her very valuable assistance in preparing this talk 2 Fairchild v Glenhaven Funeral Services Ltd [2003] 1 A.C. 32 at [45], per Lord Nicholls of Birkenhead 3 Stapleton, Cause in fact and the scope of liability for consequences, L.Q.R. 2003, 119(Jul), 388 4 Some Thoughts on Principles Governing the Governing the Law of Torts, Singapore, 19 August 2016, https://www.supremecourt.uk/docs/speech-160819-03.pdf 5 Sir Frederick Pollock, The Law of Torts: A Treatise on the Principles of Obligations arising from Civil Wrongs in the (4th ed 1895) chapter 1 6 Ibid 7 Caparo Industries plc v Dickman [1990] 2 AC 605 8 Anns v London Borough of Merton [1978] AC 728 v Stevenson9 in 1932). Caparo was, and in some quarters still is, regarded by many as finally laying down the test for determining whether a duty of care exists. To quote from the leading speech of Lord Bridge:

“in addition to the foreseeability of damage, necessary ingredients in any situation giving rise to a duty of care are that there should exist between the party owing the duty and the party to whom it is owed a relationship characterised by the law as one of ‘proximity’ or ‘neighbourhood’ and that the situation should be one in which the court considers it fair, just and reasonable that the law should impose a duty of a given scope upon the one party for the benefit of the other.”10

The first aspect of that famous three-fold test, foreseeability of damage, involves a principle; and the second aspect, proximity, could, I suppose, just about be said to do so, although it involves a heavy dollop of policy; but the third aspect, fairness justness and reasonableness, it seems to me, is pure policy, and it may be said to subsume the first two ingredients in any event.

4. The “fair, just and reasonable” test was adopted and applied by the UK Supreme Court in the so-called Christian Brothers case in 201211 in relation to the question of whether it was right to impose liability on a defendant for the tort of a third party on the ground that the defendant’s alleged vicarious liability for the third party’s actions. This approach was specifically justified by reference to policy12. Two years ago, in the Michael case13, Lord Toulson said that Lord Bridge’s speech in Caparo “was not intended to be” seen as “a blueprint for deciding cases”14. But Michael may be seen as endorsing an even more pragmatic approach, as Lord Toulson said that:

“The development of the law of negligence has been by an incremental process rather than giant steps. The established method of the court involves examining the decided cases to see how far the law has gone and where it has refrained from going. From that analysis it looks to see

9 Donoghue v Stevenson [1932] AC 562 10Caparo, p 617, per Lord Bridge of Harwich 11 The Catholic Child Welfare Society v Various Claimants & The Institute of the Brothers of the Christian Schools [2012] UKSC 56, [2013] 2 AC 1, paras 34, 47 and 94 12 ibid, para 35 13 Michael v Chief Constable of South Wales Police [2015] UKSC 2, [2015] 1 AC 1732. 14 Ibid, para 106. It is interesting that Lord Kerr’s concurring speech is very much based on the reasoning in Caparo - see eg paras 150-158 and 173 whether there is an argument by analogy for extending liability to a new situation, or whether an earlier limitation is no longer logically or socially justifiable. In doing so it pays regard to the need for overall coherence. Often there will be a mixture of policy considerations to take into account.”15

5. Another recent example of the policy-based approach is the Supreme Court’s decision last year in Patel v Mirza16, where guidance was given as to how courts should deal with cases where the defendant seeks to avoid liability by raising a defence that the claimant’s case in some way rests on illegality. (The issue in the case was in fact concerned with contractual claims involving illegality, but the Court decided to give general guidance on the topic17). Rejecting the view of the minority who sought to articulate a unifying applicable principle, Lord Toulson, speaking for the majority six of the nine Justices, cited with approval18 an observation he had made in an earlier case19:

“Rather than having over-complex rules which are indiscriminate in theory but less so in practice, it is better and more honest that the court should look openly at the underlying policy factors and reach a balanced judgment in each case for reasons articulated by it”.

6. So maybe we should not look for any principles, or at least we should not look for any but the most fundamental of principles. But some people might say that in recent years the courts have done their best to throw doubt on the correctness of even the most fundamental and elementary of principles. Any law student would tell you, or perhaps I should say that any law student would have told you, that causation is a fundamental ingredient of liability in negligence – ie that in order to succeed in a negligence claim, the claimant must show, albeit only on the balance of probabilities, that the defendant’s negligence caused the plaintiff’s loss – the so-called “but for test”. Thus in the 1968 Barnett case20, three nightwatchmen who had been poisoned by a colleague were negligently sent home by a hospital doctor and they

15 Ibid, para 102 16 Patel v Mirza [2016] UKSC 42, [2016] 3 WLR 399 17 ibid, para 2 18 ibid, para 69 19 ParkingEye Ltd v Somerfield Stores Ltd [2013] QB 840, para 52 20 Barnett v Chelsea & Kensington Hospital Committee [1969] 1 Q.B. 428, 438 to 439 subsequently died. As the evidence established that they would have died even if they had been properly treated, their relatives’ negligence claim against the hospital failed.

7. However, in subsequent cases, a purist might say that the crystal clear water in the well of principle has been polluted by the beguiling influence of public policy. Six years after Barnett, in the McGhee case21, the House of Lords upheld a claim by a plaintiff who contracted dermatitis after his employer had negligently failed to provide washing facilities after he had been exposed to clouds of abrasive brick dust, even though he could not prove that he would probably not have got dermatitis if the facilities had been made available. This was justified on the basis that it was enough that the employer’s negligence had materially increased the risk of the plaintiff getting dermatitis. Lord Wilberforce justified this on the ground that, given the uncertainty, “as a matter of policy or justice … it is the creator of the risk who … must be taken to have foreseen the possibility of damage, who should bear its consequences”22. However, 15 years later in both the Wilsher 23 and theHotson24 cases, the House of Lords refused to adopt such an approach, although it did not disapprove McGhee. As a result of this, I agree with Professor Oliphant25 that there would appear to be no clear criteria for deciding when the exceptional approach in McGhee is to be applied.

8. The material increase in risk approach was eagerly grasped by the House of Lords in the 2002 Fairchild case26, which has been described by Professors McBride and Bagshaw, not without arguable justification, as a “good example of the maxim ‘the road to hell is paved with good intentions’”27. The claimant was first employed by A and then by B, each of whom failed in his duty to take reasonable care to prevent the claimant inhaling asbestos dust and the claimant subsequently developed mesothelioma. However, the claimant’s problem was that, in the light of the expert evidence, he could have contracted this ghastly disease under only one of his employers, and it was impossible to say which of the two it was. Accordingly, as he could not prove his case against A or B on the balance of probabilities, the Court of Appeal dismissed his claim. In upholding his appeal, Lord Bingham said that “[i]f the mechanical application of generally accepted rules leads to such a result, there must be room

21 McGhee v National Coal Board [1973] 1 WLR 1. 22 [1973] 1 WLR 1 at 6. 23 Wilsher v Essex Area Health Authority [1988] AC 1074, [1988] 1 All ER 871. 24 Hotson v East Berkshire Health Authority [1987] A.C. 750.

25 Oliphant, The Law of Tort 2nd ed, p 789 ch14.13 26 Fairchild v Glenhaven Funeral Services [2003] 1 AC 32 27 McBride and Bagshaw, Tort Law 4th Edition page 285. to question the appropriateness of such an approach in such a case”28 , and Lord Nicholls said that there were “sufficiently weighty” reasons “to justify depriving the defendant of the protection [of the “but for”] test29.

9. I readily accept that on the facts of that case, many people would understandably think that it would have taken a judge with a heart of stone to uphold the Court of Appeal (who apparently did have hearts of stone) and refuse the claimant the relief which he sought. However, there is a powerful case for saying that the House should have rejected the claim and left it to Parliament to change the law. Indeed, Lord Hoffmann, who was party to the decision and justified it in his judgment by reference to material increase in risk30, has subsequently suggested, reflecting many academic critics31, that the House in Fairchild created a special exception to the “but for” principle which could not be justified by reference to any general principle and depended on a distinction which had no rational factual or legal justification32.

10. The principle in Fairchild was refined by the House of Lords four years later in Barker v Corus33, where they held that A and B could not in such a case each be liable for all the damages the claimant could claim: there should be an apportionment based on the extent to which each of them had exposed the claimant to the relevant risk. But this was reversed less than three months later by section 3 of the Compensation Act 2006, which said A and B would each be liable for all the damages. The courts having taken what may be unkindly characterised as an intellectually dubious wrong turning in Fairchild and Parliament having legislated on a rushed piecemeal basis in the 2006 Act, you might have thought that the stage was set for confusion, and so it proved. In two quite difficult decisions, Sienkiewicz34 and the “Trigger” litigation35, the Supreme Court had to deal with cases where (i) one of the employer

28 [2003] 1 AC 32, para 9 29 Ibid, para 43 30 Ibid, paras 61 to 63 31 Eg Morgan, Lost Causes in the House of Lords: Fairchild v Glenhaven Funeral Services (2003) 66(2) MLR 277–284; Stapleton, Cause in fact and the scope of liability for consequences, L.Q.R. 2003, 119(Jul), p399; Stevens, Torts and Rights (2nd edition 2007, 2012 re-print) pp 148-149; J. Morgan in R. Goldberg (eds.), Perspectives on Causation (2011); Goudkamp, The insurance law legacy of Fairchild LMCLQ 2015, 4(Nov), 443-451 32 Constitutionalism and Private Law (Cambridge Freshfields Law Lecture, 28 January 2015), as described by Lord Sumption in Zurich Insurance PLC UK Branch v International Energy Group Ltd [2016] AC 509, para 128 33 [2006] 2 A.C. 572 34 Sienkiewicz v Greif (UK) Ltd [2011] 2 AC 229, 35 Durham v BAI (Run Off) Ltd [2012] 1 WLR 867 defendants had been liable for hardly any exposure, and (ii) the extent of the liability of an insurer of an employer who was held liable under Fairchild and section 3 of the 2006 Act. The courts found themselves developing special rules of law to deal with the what Lord Hodge has judicially described as “[t]he Fairchild enclave”, ie “a special rule of causation to do justice to the victims of wrongful exposure to asbestos fibres who have contracted mesothelioma as a result”36.

11. The difficulty of creating and developing this enclave on a case by case basis was highlighted in last year’s Zurich v IEG case37, where the Supreme Court split 4-3 on the issue of how to deal with a case where the employer was covered by insurance for only part of the period for which the claimant worked for him. Should the insurer be liable for a proportionate part (the Barker-like solution) or the whole with a right to recoup form the other insurers (maybe more in line with section 3)? We were effectively forced to go on developing new rules to deal with this situation. To paraphrase Sir Walter Scott, Oh! What a tangled web courts weave when first they practice to relieve38. As Lord Reed and I said in our (minority) judgment, “there can be no real doubt but if Fairchild had been decided the other way, in accordance with normal common law principles, Parliament would have intervened very promptly. That may very well have been a better solution, but it can fairly be said that that observation is made with the wisdom of hindsight”39.

12. It can be said that Fairchild involved an extension of cases such as McGhee, but the 2004 case of Chester v Afshar40 has be seen by many as a more revolutionary departure from the principle of causation. The case involved what Lord Steyn, speaking for the majority, described as “a narrow and modest departure from traditional causation principles”41, in order to justify an award of damages to a patient who suffered injury as a result of an operation, even though she had not established that she would not have had the operation if she had been warned of the risk of the injury. Lord Steyn justified this departure from fundamental principle by saying that it was “in accord with one of the most basic aspirations of the law, namely to right wrongs. Moreover, the decision … reflects the reasonable

36 Zurich Insurance PLC UK Branch v International Energy Group Ltd [2016] AC 509, para 102 37 Zurich Insurance PLC UK Branch v International Energy Group Ltd [2016] AC 509, 38 To adapt Sir Walter Scott, Marmion (1808) 39 Ibid, para 211 40 Chester v Afshar [2004] UKHL 41, [2005] 1 AC 134 41ibid, para 24 expectations of the public in contemporary society”42. One does not need to read the brief, almost contemptuous, dissenting judgments of Lord Bingham and Lord Hoffmann (both of whom, it is to be noted, were prepared to depart from the normal principle in Fairchild) to see that there is no getting away from the fact that one of the most fundamental principles of tort law, causation, is now no longer an absolute principle at all, but must yield to policy. There is, in my view, considerable force in Professor Stevens’ view that the majority decision in Chester was wrong in principle and that “[a] preferable alternative approach to that adopted by the majority in Chester would have been to … directly award to the claimant a sum of money to reflect the fact that she had been denied the opportunity to make an informed choice”43.

13. May I turn to another controversial and difficult area of causation, namely loss of a chance. It is well established that damages can be assessed on the basis that the claimant lost a chance of a benefit. The classic case is of course Chaplin v Hicks44, whose facts capture the imagination of any law student in the same way as Carlill v Carbolic Smoke Ball45 and Donoghue v Stevenson46. Disappointed beauty contest entrants, therapeutic smoke ball advertisements and snails in ginger beer bottles all stick in the mind especially when they fall to be compared with the mundane facts of most legal cases. In Chaplin, the plaintiff had contracted to be given an opportunity (namely of winning a beauty contest), and, given that the defendant organisers wrongly deprived her of that opportunity, it seems right in principle that her damages should be assessed on the basis of valuing that opportunity, ie valuing the loss of the chance of winning.

14. Loss of a chance was authoritatively imported into tort law by the House of Lords in 1973 in a case brought under the Fatal Accidents Acts, Davies v Taylor47. In that case Lord Reid laid down what appeared to be a general proposition. He said that “[w]hen the question is … whether a certain event did or did not happen – then the court must decide one way or the other. There is no question of chance or probability. Either it did or it did not happen. But the standard of civil proof is a balance of probabilities”. But, he added,

42 ibid, para 25 43 Stevens, Torts and Rights (2007, 2012 re-print), p 165. 44 [1911] 2 KB 786 45 Carlill v Carbolic Smoke Ball Company [1893] 1 QB 256 46 [1932] AC 562 47 [1974] AC 207 “[y]ou can prove that a past event happened, but you cannot prove that a future event will happen and I do not think that the law is so foolish as to suppose that you can. All that you can do is to evaluate the chance. Sometimes it is virtually 100 per cent: sometimes virtually nil. But often it is somewhere in between. And if it is somewhere in between I do not see much difference between a probability of 51 per cent and a probability of 49 per cent.”

It appears to follow from this that damages, whether in contract or tort, for what can be characterised as the loss of a chance must be assessed by reference to that chance.

15. That is what the Court of Appeal thought in the 2000 case of Langford v Hebran48, where the claimant's chances of achieving fame and fortune as a kick-boxer at various stages in his career were evaluated on a scale from 80% to 20% and damages for his loss of earnings awarded accordingly49. However, the House of Lords have rejected (or, I would suggest, they appear to have rejected) the notion that loss of a chance should apply in medical negligence cases in Gregg v Scott50. The defendant doctor had negligently failed to refer the claimant to a specialist, as a result of which there was a nine-month delay in treating the claimant’s cancer. As a result of this delay, the claimant’s prospects of being treated so as to have a ten year period of remission was decreased from 42% to 25%, and he sought damages to compensate him for this reduction in his chances of good health. Three Law Lords rejected his claim, whereas two would have allowed it. One is not helped by the fact that there are distinct differences in the reasoning of the majority three, Lord Hoffmann, Lord Phillips and Lady Hale. However, their reasoning ultimately seems to me to come to this, that such claims should be determined by the balance of probabilities and since not only his actual 25% prospect, but also his lost 42% prospect, is below 50%, the claim failed. In other words, as the claimant would have had a less than 50% chance of enjoying ten clear years even if he had been properly referred, he had suffered no loss as a matter of law. The minority, Lord Nicholls and Lord Hope, also gave slightly different reasons, but basically they considered that, as the claimant had plainly lost something valuable and measurable, he should be compensated for it.

48 [2001] PIQR, Q160 and see also Doyle v Wallace [1998] PIQR Q146 49 To quote Lord Phillips’ description in Gregg v Scott [2005] 2 AC 176, para 119, where he also refers to Doyle 50 [2005] 2 AC 176

16. Why should there be no damages for loss of a chance in such a case? Lord Hoffmann said that the balance of probabilities was the standard approach and seems to have suggested that the loss of a chance approach only arose where “the loss of a chance of gaining an advantage or avoiding a disadvantage which depends upon the independent action of another person”51, although he also suggested that the loss of a chance approach may be limited to those cases where the claimant has suffered “financial loss, where the chance can itself plausibly be characterised as an item of property”52. Lady Hale saw the issue more pragmatically, pointing out that “If the claim is characterised as loss of a chance, those with a better than evens chance would still only get a proportion of the full value of their claim.”53

17. In my view, the decision in Gregg may be capable of being explained by reference to a rather different principle, namely the principle that if you ask a silly question you get a silly answer. Why, you may wonder, was the claim made on the basis of ten years? Why not five or twenty? Well, the justification was the expert medical evidence was that patients are deemed to be “cured” once they have been in remission for 10 years (& that was central to his claim as formulated)54. However, that may be said to be neither here nor there so far as the correct legal analysis was concerned. Presumably, if the claim had been based on, say, seven years, the figures might have been say a 55% or so chance if he had been treated and say a 35% if he had not, and so the claim would have succeeded. But the hit-and-miss nature of that analysis (why ten and not five or twenty?) shows that the whole approach to the claim was arguably mistaken. What may very well be the right approach was touched on by some of their Lordships55, but the evidence was not available to support it. That approach was described by Lord Phillis as “conventional” and explained it as follows: “The court determines what the claimant's expectation of life would have been but for the injury. The court then determines what the claimant's expectation of life is having regard to the effect of the injury. The difference between the two constitutes the ‘lost years’”56. Had that been done in Gregg, then I find it hard to see why the claimant should not have got what Lady Hale referred to as “a modest claim in respect of the ‘lost years’”57.

51 Ibid, para 82 52 Ibid, para 83 53 Ibid, para 225. 54 Ibid, para 64 55 Ibid, Lord Nicholls at para 57, Lord Phillips at para 131 and Lady Hale at para 210 56 Ibid, para 131 57 Ibid para 208

18. The final point about the decision in Gregg is that it may be difficult to reconcile with the decision in Chester. As has been tellingly, if rhetorically, asked by Maskrey and Edis “[w]hy should the ‘Right’ to decide whether to accept treatment be accorded greater protection and value by the law than the logically and chronologically prior right to be told that such treatment is available and could be beneficial?”58

19. Even where the normal causation principles appear to apply, policy comes into play on issues of damages. As Lord Sumption explained earlier this year in BPE Solicitors case59, “[i]t is generally a necessary condition for the recovery of a loss that it would not have been suffered but for the breach of duty. But it is not always a sufficient condition.” He added that the extent of a defendant’s liability will often “depend on a developed judicial instinct about the nature or extent of the duty which the wrongdoer has broken”60.

20. The cases I have so far referred to tend to suggests that principle rather than policy may be a dangerous guide to those analysing, advising on, debating, or deciding cases involving claims in tort. However, clarity and predictability are vitally important ingredients of the rule of law. Accordingly, it can be said with considerable force that it is quite right for the judges to be striving to identify principles in the field of tort law. Even if a particular principle cannot always apply, it may prove reliable in the great majority of cases. But there is no getting away from the fact that there are real risks in developing principles in the field of torts, as they may not infrequently operate to mislead rather than to help.

21. There is a strong argument that, in some areas at least, it may be more helpful to abandon principle and to take a stand on policy. Of course, policy may often be less reliable than a principle at indicating a specific outcome in a specific case. But policy may often be more reliable in the sense that it is less prone to founder on exceptions or turn out to be unsound. And, since principle is based on policy, there can be said to be less of a risk of losing the thread if one applies principle rather than policy. Certainly, there is much to be said for the view that considerable caution should be employed before a court adopts or approves a

58 Maskrey and Edis, Chester v Afshar and Gregg v Scott: mixed messages for lawyers (2005) JPIL 205 at p 222. 59 BPE Solicitors v Hughes-Holland [2017] 2 WLR 1029, para 20 60 Ibid principle on the basis that it is susceptible of general application – and even more caution if it is suggested that the principle should not be subject to exceptions or development.

22. There are of course arguments, well summarised by Professor Stevens61, as to why judges should leave policy alone: judges are unelected, judges lack the technical competence, and policy leads to less certain outcomes. It is often said that judges are often neither informed or experienced enough to make the economic and social assessments which questions of policy in the field of torts so often involve. I wonder. Many of the most important judicial decisions in the field of torts seem to me to involve those sort of assessments. Donoghue v Stevenson62 is a prime example: it involved a considerable extension of product liability; Hedley Byrne63 involved extending liability for negligent misstatements to consequential economic loss; the same is true of SAAMCo64, the surveyor’s negligence “cap” case considered in BPE Solicitors discussed earlier. And, if one casts one’s eyes more widely, perhaps the most marked development in the common law in the past fifty years has been the very substantial growth in domestic judicial review: nobody can deny that it is a development for which the judges are responsible, or that it is a development with substantial macro-social and economic implications.

23. More broadly, the life of the common law is, famously, experience not logic65. Although any particular judge may have limited experience, one of the strengths of the common law is that a judge’s decision in a particular case is ultimately based on centuries of judicial experience of multifarious cases. And, particularly in a field such as tort law, any principles which have been developed have been initiated and developed by judges, and those principles are inevitably very much based on policy. Accordingly, one may wonder how much there is in the notion that judges are not competent to deal with, or to decide cases on the basis of, policy issues.

24. Of course, there are different sorts of policy issues, and in some cases, it would be right for a judge to say that the question whether the law should be extended in a certain direction is not a point which should be determined by the judiciary but by the legislature. The very fact

61 Robert Stevens, Torts and Rights (2007, 2012 reprint) pp 308-310 62 [1932] AC 562 63 Hedley Byrne v Heller & Partners Ltd [1964] AC 465 64 South Australia Asset Management Corpn v York Montague Ltd [1997] AC 191 65 Oliver Wendell Holmes, The Common Law (1881) p 1 that judges have the right, indeed sometimes the duty, to develop the law means that self- restraint is an important weapon in the judicial armoury: it is important for a judge to know when to stand back and leave policy-based developments in the law to the legislature. But that is a very different thing from saying that judges should not develop the law themselves by reference to policy.

25. Quite apart from this, if a judge makes a policy-based decision with which the legislature is not happy, the remedy in a system with parliamentary supremacy, such as we enjoy in the UK, lies with Parliament. Any decision made by a court can always be reversed by the legislature.

26. On the other hand, the notion that Parliament can be reliably expected to step in is often little more than a pious hope, given the enormous pressure on legislative time and the understandably slender political importance which some may think is attributed to most legal issues (as is apparent from Parliament’s failure to take up judicial invitations, indeed judicial pleas, to enact a law on illegality, leading the courts to do so, in Patel v Mirza).

27. The fact is that, true to form, the common law wants to have the best of both worlds. It develops and applies principles as far as it can in order to introduce predictability and logical cohesion into the law, but it accepts that, with the infinite variety of human experiences, and with developments in the social, economic, ethical and technological spheres, hardly any principles can be applied blindly and most principles will have to be subject to exceptions, to changes or even to discarding at various stages of their existence.

28. And, sometimes at least, practising lawyers must wonder whether judges just want to ensure that legal advisers are kept busy or, as some might see it, kept confused.

David Neuberger

13 May 2017